But we must not on that account 'empty out the child with the bath,' and conclude that there is no such thing as a 'biogenetic law' or recapitulation of the phylogeny in the ontogeny. Not only is there such a recapitulation, but—as F. Müller and Haeckel have already said—ontogeny is nothing but a recapitulation of the phylogeny, only with innumerable subtractions and interpolations, additions and displacements of the organ-stages both in time and place. It would be a great mistake to conclude from the fact of these manifold alterations that the whole proposition of the recapitulation of the phylogeny in the ontogeny is erroneous, or at least valueless. If its only use were to enable us to read the racial history of a species out of its germinal history, it is intelligible enough that we might be led to give it up in despair, but I think that the main thing is to get some insight into the history of the ontogeny, and there can be no doubt that this can have been built up on no other foundation than upon the racial history. What is new could only have arisen from what was already in existence, and everything in ontogeny, not only the palingenetic stages which still represent in some measure the facies of fully-formed ancestral stages, but also the cœnogenetic stages, like the pupa-stage we have already discussed, have arisen historically, nothing de novo, but all in connexion with what was already present. But what was first present was in all cases the stages of the ancestral forms.
It is undoubtedly of the greatest value to be able to penetrate more and more deeply into embryonic development, and to discover more precisely the changes that have taken place throughout its course in the originally existing material of ancestral forms. But it must not be forgotten that, all transformations notwithstanding, so much of the racial history is still very plainly indicated in the germinal history, that this must always remain for us a most important source from which to draw conclusions in regard to the phyletic development of any animal group. I admit that these conclusions have sometimes been drawn with too great confidence, but even if we cannot regard as well founded Haeckel's view that in the ontogeny of Man there are fourteen different ancestral stages recognizable, a protist stage, a gastræa stage, a prochordate, an acranial, a cyclostome, a fish-stage, and so on, we must recognize that the unicellular stage of ontogeny, with which even now the development of every human being begins, undoubtedly repeats the facies of an ancestor, although greatly altered; for we must be descended from unicellular organisms. The essential part of this ancestral stage is thus preserved in the ontogeny, and only what is special and in some measure due to chance, that is, to adaptation to special conditions of existence, has been modified.
It has been supposed that the proposition that phylogeny is recapitulated in the ontogeny is disproved, because the ontogenetic stage must always contain within it the primordia of the later stages which have been added since the corresponding phylogenetic stage. It is certain that the egg-cell or the sperm-cell of Man contains, though in a form not recognizable by us, all the determinants of the perfect human body, but this neither affects its nature as a cell nor its particular form as ovum or spermatozoon. It is essentials that are important in this comparison, not accessories. Neither can I agree with Hensen's argument when he says that the 'recapitulation-idea' is erroneous, because the actual course of ontogeny is the 'best and only possible one,' which, apart from previous history altogether, must of necessity be followed. Certainly the actual course is the best, and under the given circumstances the only possible one, but that does not exclude recapitulation, on the contrary it implies it, for ontogeny could at no time have arisen from a tabula rasa, but only from what was historically existent.
I do not propose to examine each of Haeckel's ancestral stages in Man's pedigree, or to estimate the degree of probability with which they may be deduced from the ontogeny; but that Man's ancestry does, in a general way, include such a series of phyletic stages may be admitted, even if we grant that many of these stages are now no longer represented in the ontogeny as stages of the developing organism as a whole, but only by stages of individual organs or group of organs. Thus it may be disputed whether there is still a fish-stage in human development, but it cannot be disputed that the rudiments of 'gill-arches' and 'gill-clefts,' which are peculiar to one stage of human ontogeny, give us every ground for concluding that we possessed fish-like ancestors.
As we now know that the history of a given mode of embryogenesis has involved numerous time-displacements of the organ-rudiments, we must attach all the more weight to the developmental history of the individual parts and characters, in which the phylogeny can often be read more clearly than in the stages of the organism as a whole, and we can probably find out important laws in this way.
As far back as 1873 Würtemberger investigated the fossil ammonites with special reference to this point. He was concerned even more at that time with finding proofs of the theory of descent in general, and this was the first case in which any one succeeded in demonstrating phyletic transformation-series of species, deposited one above the other in a corresponding series of geological strata, and connected by transition forms lying between these. In studying this interesting material, of which many examples were at his disposal, Würtemberger proved that the variations which had taken place in the spirally coiled shell in the course of ages appeared first on the last whorl, and then subsequently extended to the one before this, and thence to the still younger whorls of the shell. Meanwhile the last whorl not infrequently exhibited another new character. Thus, for instance, protuberances on the shell were shifted in the course of the phylogeny from the last convolution to the second last, and later to the third last, and so on, while at the same time the last convolution showed the protuberance changed into spines. In other words, the new phyletic acquirements first appeared in the mature animal (in the last-formed whorl or chamber of the shell), but were subsequently shifted back in the ontogeny to younger stages in proportion as new transformations of the mature animal appeared. Thus there was, so to speak, a retraction of the phyletic acquisitions of the mature animal deeper and deeper into the germinal history of the species.
About the same time—in the seventies—I obtained similar results from living species when I was attempting to work out the ontogeny of the markings on the external skin of the caterpillars of certain butterflies, and I should like to submit a short account of these.
In one of the early lectures we discussed the protective and defensive colours of caterpillars in general, and those of caterpillars of the Sphingidæ in particular. I showed that those naked caterpillars which live on plants among the grass, or on the grass itself, are often not only green, like fresh grass-stalks, or yellowish-grey, like dry ones, but all the larger forms also exhibit light, usually white, longitudinal lines, which, by mimicking the sharp light reflections on the grass-stems, heighten the protective resemblance.
We also spoke of the light transverse stripes, often marked with pink or lilac-blue, of many of the large green caterpillars which live on trees and bushes, and whose likeness to the leaves is heightened by this imitation of the lateral veining of a leaf; and finally we mentioned the warning coloration indicative of unpleasant or nauseous taste, among which must be classed not only vivid contrasts of colour, but also specially conspicuous elements of colour such as light ring-spots upon a dark ground. These different colour schemes which protect the caterpillars from their enemies are usually only to be found in the adolescent caterpillar, not in the very small one which has just emerged from the egg, and the development of the markings in the individual life clearly shows that the phylogeny of the markings is more or less obviously contained in the ontogeny.
There are three different schemes of marking which occur in the caterpillars of hawk-moths or Sphingidæ—longitudinal striping, obliquely transverse striping, and spots. Longitudinal striping pure and unmixed is now found only in a few species, for instance in the caterpillar of the Macroglossa stellatarum (Fig. 115), in which a white longitudinal line, beginning at the tip of the tail, runs up each side of the body to the head as a 'sub-dorsal stripe' (sbd). These, with other two similar stripes, effectively secure the fairly large caterpillar from discovery when it is among grass and herbs.